Comparison of Farmer-Participatory Research Methodologies

H A Freeman

International Crops Research Institute for the Semi-Arid Tropics

ICRISATInternational Crops Research Institute for the Semi-Arid Tropics

PO Box 39063, Nairobi, Kenya

1030–2001

Working Paper Series no. 10

Socioeconomics and Policy Program

Comparison of Farmer-ParticipatoryResearch Methodologies:

Case Studies in Malawi and Zimbabwe

About ICRISAT

The semi-arid tropics (SAT) encompasses parts of 48 developing countries including most of India,parts of southeast Asia, a swathe across sub-Saharan Africa, much of southern and eastern Africa, andparts of Latin America. Many of these countries are among the poorest in the world. Approximatelyone-sixth of the world’s population lives in the SAT, which is typified by unpredictable weather,limited and erratic rainfall, and nutrient-poor soils.

ICRISAT’s mandate crops are sorghum, pearl millet, finger millet, chickpea, pigeonpea, andgroundnut; these six crops are vital to life for the ever-increasing populations of the semi-arid tropics.ICRISAT’s mission is to conduct research which can lead to enhanced sustainable production ofthese crops and to improved management of the limited natural resources of the SAT. ICRISATcommunicates information on technologies as they are developed through workshops, networks,training, library services, and publishing.

ICRISAT was established in 1972. It is one of 16 nonprofit, research and training centers fundedthrough the Consultative Group on International Agricultural Research (CGIAR). The CGIAR is aninformal association of approximately 50 public and private sector donors; it is co-sponsored by theFood and Agriculture Organization of the United Nations (FAO), the United Nations DevelopmentProgramme (UNDP), the United Nations Environment Programme (UNEP), and the World Bank.

AbstractFarmer-participatory research (FPR) can help improve the effectiveness of technology development, raiseadoption rates, and increase the payoff to agricultural research. However, there is wide diversity of opinion aboutthe scope and nature of farmer participation in the implementation of FPR. This study compared different FPRapproaches being used by various organizations at six on-farm sites in Malawi and Zimbabwe, with a view todeveloping guidelines for future work, particularly in developing technologies targeted at women farmers. Thestudy found differences between the various approaches in diagnostic activities, planning of experiments,assessment of results, and strengthening of farmers' capacity to conduct their own experiments. Overall, however,the various methods were not completely distinct, but represented a continuum from traditional research-led, toresearcher-led with farmer input, to farmer-led with research input. The results indicated that when there are noclear procedures to directly target women farmers they tend to be under-represented, and gender issues are notsufficiently integrated into the research process. PRA techniques were also found to be useful in planningexperiments, but require considerable investment of time by researchers and extension staff.

Researchers therefore need to formulate clear objectives for conducting different types ofon-farm experiments, ensure that farmers understand their roles in each case, and finally integrate different typesof research to develop technologies that are practical and profitable.

RésuméComparaison des méthodologies de recherche menée avec la participation des paysans : Etudes de cas auMalawi et au Zimbabwe. Les recherches menées avec la participation des paysans (FPR) peuvent améliorer lamise au point des technologies, augmenter le taux d’adoption et accroître les retombées positives de la rechercheagricole. Il existe cependant, une large diversité d’opinions sur la portée et la nature de la participation despaysans à la mise en œuvre de la recherche participative. La présente étude a comparé les différentes approchesen matière de recherche participative adoptées par différentes organisations dans six sites en milieu paysan auMalawi et au Zimbabwe, en vue d’élaborer des directives pour les travaux futurs, en particulier pour la mise aupoint de technologies destinées aux femmes en milieu rural. L’étude a trouvé qu’il y a des différences entre lesdiverses approches adoptées dans les activités de diagnostic, de planification des expériences, d’évaluation desrésultats et de renforcement de la capacité des paysans à réaliser leurs propres expériences. En général, cesdifférentes méthodes ne se distinguaient pas totalement les unes des autres, mais formaient un continuum qui vades méthodes de recherche sans participation des paysans, ensuite les recherches faites par les chercheurs avec laparticipation des paysans et à la fin la recherche faite par les paysans en collaboration avec les chercheurs. Lesrésultats de l’étude indiquent que lorsqu’il n’y a pas de procédures claires permettant de cibler directement lesfemmes productrices, ces dernières ont tendance à être sous-représentées et les questions de genre ne sont passuffisamment prises en compte dans le processus de recherche. L’étude a également trouvé que les techniques deFPR sont utiles pour la planification des expériences, mais exigent un important investissement de temps de lapart des chercheurs et du personnel de vulgarisation. Les chercheurs doivent donc définir des objectifs clairs pourla conduite de différents types d’expériences en milieu paysan; s’assurer que les paysans comprennent leur rôledans chaque cas et combiner différents types de recherche pour mettre au point des technologies pratiques etrentables.

Citation:Citation:Citation:Citation:Citation: Freeman, H.A. 2001. Comparison of farmer-participatory research methodologies: case studies inMalawi and Zimbabwe. Working Paper Series no. 10. PO Box 39063, Nairobi, Kenya: Socioeconomics andPolicy Program, International Crops Research Institute for the Semi-Arid Tropics. 28 pp.

Titles in the Working Paper Series aim to disseminate information and stimulate feedback from the scientificcommunity. A large number of similar as well as informal publications have been distributed by ICRISAT’sSocioeconomics and Policy Program, under various titles: Progress Reports, Discussion Papers, and OccasionalPapers.

Comparison of Farmer-Participatory ResearchMethodologies: Case Studies in

Malawi and Zimbabwe

H A Freeman



2001

About the author

HAHAHAHAHA Fr Fr Fr Fr Freemaneemaneemaneemaneeman Senior Scientist (Economics), ICRISAT, PO Box 39063, Nairobi, Kenya

© 2001 by the International Crops Research Institute for the Semi-Arid Tropics

All rights reserved. Except for quotations of short passages for the purposes of criticism or review, no part of this publicationmay be reproduced, stored in retrieval systems, or transmitted in any form or by any means, electronic, mechanical,photocopying, recording or otherwise, without prior permission from ICRISAT. It is hoped this copyright declaration will notdiminish the bonafide use of research findings in agricultural research and development.

The views expressed are those of the author and not necessarily those of ICRISAT. The designations employed and thepresentation of material in this publication do not imply the expression of any opinion whatsoever on the part of ICRISATconcerning the legal status of any country, territory, city, or area, or of its authorities, or concerning the delimitation of itsfrontiers or boundaries. Where trade names are used, this does not constitute endorsement of or discrimination against anyproduct by the Institute.

This publication is part of the research project “Will women farmers invest in improving their soilfertility management? Participatory experimentation in a risky environment”, funded by the

Department for International Development (DFID), UKDepartment for International Development (DFID), UKDepartment for International Development (DFID), UKDepartment for International Development (DFID), UKDepartment for International Development (DFID), UK.

Contents

Introduction ....................................................................................................................................... 1

Methodology...................................................................................................................................... 1

FPR methods...................................................................................................................................... 3

Selecting farmers for participation in on-farm trials ......................................................................... 3

Farmer participation in technology development and testing: evidencefrom the case studies ......................................................................................................................... 4

– Traditional research-led approach .............................................................................................. 4

– Researcher-led with farmer input ............................................................................................... 6

– Farmer-led with researcher input ............................................................................................... 10

Guidelines for FPR: key issues.......................................................................................................... 12

Conclusions ....................................................................................................................................... 14

Acknowledgments ............................................................................................................................. 15

References ......................................................................................................................................... 15

Annex 1. Case study sites, FPR methodology, and lead organization(s) .......................................... 16

Annex 2. Checklist for discussion with farmers ................................................................................ 16

Annex 3. Types of trial implemented at case study sites ................................................................... 18

Acronyms

AGRITEX Department of Agricultural, Technical and Extension Services (Zimbabwe)

CIMMYT Centro Internacional de Mejoramiento de Maïz y del Trigo

DFID Department for International Development (UK)

DR&SS Department of Research and Specialist Services (Zimbabwe)

ICRISAT International Crops Research Institute for the Semi-Arid Tropics

NGO Nongovernmental organization

PRA Participatory Rapid Appraisal

SADC Southern African Development Community

SMIP Sorghum and Millet Improvement Program

TSBF Tropical Soil Biology and Fertility Program

1

IntroductionIntroductionIntroductionIntroductionIntroductionFarmer-participatory research (FPR) describes a process for developing and testing agriculturaltechnologies. This approach encompasses diverse research and research-related activities that rangefrom informal surveys with a few farmers, to conducting research with farmer involvement, tocommunity empowerment, to technology development and dissemination by extension services andother development institutions. In principle, however, FPR describes a process that is based on dialogbetween farmers and researchers in order to develop improved technologies that are practical,effective, profitable, and will solve identified agricultural production constraints.

Few would question the need for farmer participation in the research process. It is widely viewedthat incorporating farmers' priorities and assessment can improve the effectiveness of technologydevelopment, raise adoption rates, and increase the payoff to agricultural research. However, inpractice there is wide diversity of opinion about the scope and nature of farmer participation in theimplementation of FPR.

The DFID-supported project implemented by ICRISAT, Will women farmers invest in improvingtheir soil fertility management? Participatory experimentation in a risky environment, seeks todevelop research methods and practical soil fertility management options for smallholder farmers,particularly women farmers who are heads of households (Snapp 1999). A key component of thisresearch is to provide guidelines on FPR methods that will facilitate the integration of farmers'assessment of technology options into the development, testing, and dissemination of practical soilmanagement options1. To accomplish this the project tested one FPR method at each of six case studysites, three each in Malawi and Zimbabwe. The various methods were all FPR, but involved varyingdegrees of farmer participation: traditional research-led, research-led with farmer input, and farmer-led with research input. The goal was to compare the effectiveness of different methodologies indeveloping and testing soil management technologies for resource-poor farmers, particularly women.The objective of this study is to assess researchers' characterization of various alternative FPRmethods, identify the main differences between them, and thus compare the different approaches. It isexpected that the study will provide guidelines for FPR, particularly as it relates to developingtechnologies targeted at women.

MethodologyMethodologyMethodologyMethodologyMethodologyA case study approach was used. Informal interviews were conducted with site leaders (scientists orstaff of collaborating national research programs or NGOs) as well as the technicians and enumeratorswho supervised and monitored the trials at the case study sites. These informal interviews weresupplemented with information from project records and two workshops organized by the project.Interviews were conducted with individual farmers and groups of farmers during April and May 2000at five villages in Malawi and six villages in Zimbabwe (Fig.1, Annex 1). A checklist was used in theinterviews to ensure that all major topics were covered (Annex 2).

1. The FPR research at two sites in Zimbabwe was undertaken under the regional SADC/ICRISAT Sorghum and Millet ImprovementProgram (SMIP).

2

FigurFigurFigurFigurFigure 1. Location of field sites in Malawi and Zimbabwe.e 1. Location of field sites in Malawi and Zimbabwe.e 1. Location of field sites in Malawi and Zimbabwe.e 1. Location of field sites in Malawi and Zimbabwe.e 1. Location of field sites in Malawi and Zimbabwe.

3

FPR MethodsThe primary aim of this study is to compare different FPR methods. The research implemented at thecase study sites provides an opportunity to evaluate whether the various methods were indeedsignificantly different, and the implications for technology development and testing, particularly as itrelates to women farmers. The study therefore focuses on technology adoption and adaptation ratherthan the technologies per se. The framework for comparison uses the stages in the research processand the level of farmer/researcher participation in decision making. As is commonly done in the FPRliterature (eg Lilja and Ashby 2000, Okali et al. 1994), we divided the research process into threestages: diagnostic; planning of experimentation; testing and adapting. In the diagnostic stage theresearch problem is diagnosed, opportunities for research identified, priorities set, and potentialsolutions identified. In the experiment-planning stage the potential options selected forexperimentation are evaluated, and decisions made about who conducts the experiment, where andhow, and with what inputs. The testing and adapting stage involves implementation and monitoring oftrials, data collection, and assessment of results.

Selecting Farmers for Participation in On-Farm TSelecting Farmers for Participation in On-Farm TSelecting Farmers for Participation in On-Farm TSelecting Farmers for Participation in On-Farm TSelecting Farmers for Participation in On-Farm TrialsrialsrialsrialsrialsFarming communities in southern Africa are not homogeneous. Farmers differ in social status, wealth,access to and control over resources, and proclivity to conduct research. Research activities thereforefocused on the needs of low-resource farmers, particularly women farmers (Snapp 1999). Gender wasidentified as a key variable in the participatory testing and dissemination of technologies. The casestudies, however, suggest that in practice there were few attempts to directly target and select womenfarmers for on-farm trials in any of the FPR approaches. This was due partly to the lack of proceduresfor selecting women farmers. Because the project did not set standard procedures for integratinggender concerns into the participatory testing of technologies, each research team developed its ownstrategy for selecting participants for experimentation.

There were four main strategies for selecting farmers to participate in on-farm trials. In part, eachstrategy reflected the research team's perception of farmer participation. First, a strategy that wasfrequently used in the mother-baby trials was to select village headmen to host the mother trials, withother farmers volunteering to participate in the baby trials. The rationale was that farmers frequentlymeet at the headman's compound for various community and development initiatives. Thus, locatingthe trials in the headman's field offered easy access to the mother trial, facilitating farmers' observationand comparison of the mother trial with their own baby trials. However, all the village headmen at thecase study sites were influential and relatively wealthy male farmers, not representative of resource-poor farmers or women farmers. Because they were selected solely for being local leaders they did notbring any specific expertise into the experimentation process.

The second strategy was ask for volunteers during village meetings, and from them select farmersto participate in on-farm trials. This strategy was used by the SADC/ICRISAT Sorghum and MilletImprovement Program (SMIP) and the Tropical Soil Biology and Fertility (TSBF) project inZimbabwe, and in ICRISAT's baby trials in Malawi and Zimbabwe. The strength of this approach wasthat farmers were selected based on their willingness and ability to host the trials. But it is alsoplausible that male and/or resource-rich farmers who could afford the resources were most likely tovolunteer and be selected. This perception was confirmed during field work. Many participants in thebaby trials and the farmer-led trials were relatively better-off males. However, at the SMIP projectsites the majority of farmers attending meetings and volunteering to host trials were women (GHeinrich, personal communication). In some cases researchers considered socioeconomic differences

4

– including gender – among farmers volunteering to host the trials, but it was not clear whethersystematic consideration was given to the selection of women. Thus, the selected farmers may or maynot be representative of women farmers.

A third strategy used by the Centro Internacional de Mejoramiento de Maïz y del Trigo(CIMMYT) in Zimbabwe and in baby trials in Malawi, was to let farmers select participants withintheir communities. These farmers were selected because of the perception that they wereknowledgeable about farming, willing to share knowledge and experiences, and hence best placed torepresent their communities. An implicit assumption in several cases was that these farmers wouldautomatically share their knowledge with researchers and other farmers in their community. In at leastone case the issue of selected farmers sharing their knowledge and experience with others was raisedduring the selection process. In this case, after the farmers had been selected the make up of the groupwas compared with a community wealth ranking exercise conducted earlier, and the groupcomposition was adjusted accordingly.

In the fourth strategy, used by CIMMYT in Malawi, farmers were selected by extension staff andenumerators. Here also there was high probability that selection would be biased against womenbecause extension staff and enumerators were inclined to select better endowed farmers or thoseperceived to be more innovative or research minded. In several cases these criteria tended to favormale farmers. One notable exception was in Malawi, where CIMMYT researchers used informationfrom a diagnostic baseline survey to select women farmers for the baby trials.

Farmer Participation in TFarmer Participation in TFarmer Participation in TFarmer Participation in TFarmer Participation in Technology Development and Technology Development and Technology Development and Technology Development and Technology Development and Testing:esting:esting:esting:esting:Evidence from the Case StudiesEvidence from the Case StudiesEvidence from the Case StudiesEvidence from the Case StudiesEvidence from the Case StudiesCase study sites were located in and around villages where on-farm trials were being implemented aspart of another research program. A full comparison involved a control village where there were nointerventions, a village where interventions were limited to demonstration plots, a researcher-led trialvillage, and a farmer-led trial village. Researchers characterized the approach used at each site astraditional research-led, research-led with farmer input, and farmer-led with or without research input.It was hypothesized that the largest degree of researcher control would be in a traditional trialapproach, with control diminishing in favor of farmer control as the research evolved towards a farmerempowerment approach.

TTTTTraditional research-led approachraditional research-led approachraditional research-led approachraditional research-led approachraditional research-led approach

Traditional research-led approaches were implemented in Malawi and Zimbabwe. These trials aremeant to demonstrate ‘best bet’ technology options defined by researchers. The trials are designed andmanaged by researchers with inputs from extension, using a standard trial design and approach. Fielddays are held and researchers may organize farmer exchange visits to other demonstration trials.

Diagnosis stageDiagnosis stageDiagnosis stageDiagnosis stageDiagnosis stage

No baseline diagnostic survey or participatory problem analysis was conducted at the case study sitesto help identify major constraints and their causes and effects, before trials were implemented.However, researchers had access to an earlier baseline diagnostic survey which provided informationon local socioeconomic conditions. Researchers had developed a menu of best-bet soil fertilitytechnology options for Malawi and Zimbabwe based on their potential acceptability to smallholderfarmers. The specific options tested in the on-farm trials were selected by researchers and presented to

5

farmers as potential solutions for addressing the perceived productivity problem. Research andextension staff organized village meetings to consult with farmers. At these meetings researchersdiscussed trial plans and their implementation with farmers and selected farmers for the trials.Researchers initiated and led the discussions. Farmers responded to enquiries but it was apparent thattheir views and opinions were not actively sought in identifying technology options forexperimentation.

Planning and implementation of trialsPlanning and implementation of trialsPlanning and implementation of trialsPlanning and implementation of trialsPlanning and implementation of trials

The trials were formal experiments designed and implemented by researchers on farmers' fields, usinga traditional experimental design with randomized experiments and replicates. Farmers provided theland and, in some cases, labor for plowing and weeding as laid out in the trial plan. Researchersprovided all other inputs, and made all decisions regarding quantity, timing, and method of applicationof inputs. All activities were carried out either by research staff or by labor hired by researchers.

The level of interaction between scientists and farmers was low. Scientists were present at theinitial consultative village meeting. They also visited the trials to observe the performance of theexperiments. Technicians were also present at the village meeting, and managed and supervised trialactivities. Enumerators who resided in the area monitored the trials on farmers' fields. Data werecollected on traditional biophysical parameters such as method and date of planting, crop stand, weedcoverage, dates of weeding, and rainfall. The data were passed on to researchers for analysis anddissemination.

The high level of researcher control in these trials was expected to lead to the collection of highquality biophysical data. However, logistics problems in managing and supervising the trials made thisdifficult. Enumerators monitored the trials but there was little contact with scientists and technicianswho had authority to make and change decisions. For example, in Zimbabwe, research officers arebased in Harare, approximately 300 km from the trials in Zimuto. Visits to research-led trials anddiscussions with farmers hosting these trials in both Malawi and Zimbabwe suggested significantvariation in the level of supervision and management of trial activities, which affected the quality ofthe trials and the data collected.

Assessing results from on-farm trialsAssessing results from on-farm trialsAssessing results from on-farm trialsAssessing results from on-farm trialsAssessing results from on-farm trials

Farmer assessment of the technologies was limited to field days organized by research and extensionstaff2. These field days were meant to demonstrate the potential of the technology options to farmers inthe area. Researchers led discussions and answered farmers' questions. Ideally, farmers attending thefield days would assess the trials and provide ideas for experimentation. But many farmers, includingthose hosting the trials, felt it was difficult for them to realistically assess the trials because of theirlimited involvement. However, some farmers who hosted trials expressed a desire to try out, on theirown, some of the treatments that looked promising.

TTTTTraining farmers, disseminating informationraining farmers, disseminating informationraining farmers, disseminating informationraining farmers, disseminating informationraining farmers, disseminating information

Technicians and enumerators were responsible for collecting data and monitoring the trials. Resultswere analyzed by researchers and disseminated to typical research audiences. Thus, the experimentsdid not strengthen farmers' capacity to conduct their own experiments.

2. Project staff reported visiting trial sites and communicating the results back to farmers after this survey was conducted.

6

Research-led with farmer inputResearch-led with farmer inputResearch-led with farmer inputResearch-led with farmer inputResearch-led with farmer input

This method was research-led but with greater farmer involvement in testing and assessingtechnologies. Activities included both research-managed and farmer-managed trials and wereimplemented jointly by ICRISAT, CIMMYT, TSBF, and national research and training organizationsin Malawi and Zimbabwe. Because different organizations and locations were involved, the exactapproach varied, but in general they could all be characterized as research-led with farmer input.

One method used was the mother-baby trial design (Snapp 1999). This approach seeks to facilitatecommunication between researchers and farmers by incorporating farmers' input early in the researchprocess. The mother trials are researcher designed and managed, fully replicated on-farm trialsinvolving a range of technology options. The baby trials are a subset of the technologies tested in themother trial, with which the host farmers experiment. The baby trial can be equivalent to one fullreplicate of the mother trial, or only a partial subset. For statistical purposes each host farmer isconsidered as a replicate. This design attempts to meet researchers' objectives to generate reliablebiophysical data and simultaneously allow farmers to compare a subset of technologies in the babytrials with the wider range in the mother trial. Both ICRISAT and CIMMYT implemented mother-baby trials in Malawi and Zimbabwe, but there were differences in the level of farmer participation atdifferent stages in the research process. These differences mainly reflected researchers' diverse viewsof how FPR should be implemented.

SMIP and TSBF used a different approach to research-led experimentation with farmer input inZimbabwe. Researcher-managed and farmer-managed trials were conducted simultaneously onfarmers' fields. The researcher-managed trials served as focal points for field days and providedfarmers with opportunities to observe and compare different technologies. In the SMIP trials farmersselected the technology options they wished to test, after discussions with researchers. Consequently,researchers and farmers developed the trial design together. In both cases farmers conducted their ownexperimentation using their desired input levels and management practices.


The institutions involved in the research used different approaches to define the research problems andtheir causes.

Mother-baby approach. The DR&SS/ICRISAT/CIMMYT mother-baby trials in Zimbabwe did notconduct reconnaissance or diagnostic baseline surveys to help guide the research, but usedsocioeconomic data from earlier surveys in the case study villages (personal communication, KVaughan). In Malawi researchers used information from an earlier participatory problem analysis,informal farmer consultations in village meetings, and data from a crop management reconnaissancesurvey to help identify the major productivity constraints and plan research activities (personalcommunication, S Snapp). In addition, baseline surveys were conducted at the Malawi sites to obtaindetailed information on farmers' socioeconomic characteristics, crop management practices, andproductivity constraints. However, the results were not readily available when the first year's researchactivities were being planned. Notwithstanding this, farmer participation in problem diagnosis waslimited to providing answers to questions determined by researchers. To this extent, there was onlyone-way communication between farmers and researchers.

SMIP/TSBF approach. In order to identify opportunities for experimentation, information onproductivity constraints at the trial sites was obtained from two sources: baseline diagnostic surveysand village meetings. The baseline surveys were carried out by SMIP specifically to diagnose

7

productivity constraints and identify opportunities for experimentation. The village meetings wereorganized by researchers and extension staff, with the same objectives. Farmers identified severalconstraints during these meetings. SMIP and TSBF conducted separate trials at different locations.One TSBF location had been earlier surveyed by SMIP, so TSBF was able to use the survey data toplan the experiments.

Selecting and prioritizing technology options for Selecting and prioritizing technology options for Selecting and prioritizing technology options for Selecting and prioritizing technology options for Selecting and prioritizing technology options for experimentationexperimentationexperimentationexperimentationexperimentation

Mother-baby approach. At the village meetings in Malawi and Zimbabwe, research and extension staffinformed farmers about what technology options were likely to resolve the perceived productivityproblems. Researchers explained the trials that were planned and the potential impact of thetechnology options selected for experimentation. All treatments in the mother-baby trial designs wereresearcher-designed best-bet options that were developed to address specific production constraintshypothesized by researchers. Although farmers participated in village meetings it was apparent fromthe set experiments conducted that they did not have specific inputs in choosing what technologies totest.

SMIP/TSBF approach. In the SMIP trials in Zimbabwe, priorities for experimentation weredetermined based on farmers' assessment of constraints, i.e. information from baseline surveys andvillage meetings. At these meetings farmers suggested options for experimentation based on theirexperience and prior knowledge from an NGO-sponsored exchange visit to other experimental sites.Researchers also suggested technology options that had worked well in other areas. Throughnegotiation, researchers and farmers eventually agreed on what options would be tested.

The TSBF trials in Zimbabwe also used village meetings to set priorities for experimentation.Farmers raised concerns about some of the technology options that researchers suggested. Theseconcerns were discussed, and this led to a set of research issues which provided a basis for selectingtechnology options for testing. Final selection, for both researcher-managed and farmer-managedexperiments, was done by researchers. The researcher-managed trials were a combination ofexperiments that had performed well under research management in other areas, and a limited set ofexperiments to address research issues identified jointly at the village meeting. The farmer-managedtrials were a combination of technologies suggested by researchers, and those which farmers wereinterested in trying on their own in the light of research issues identified at the village meetings.

Planning stagePlanning stagePlanning stagePlanning stagePlanning stage

Mother-baby approach. Researchers provided information on the mother/baby trial design and thetechnology options at the village meetings. The mother trials were designed and managed byresearchers using traditional experimental approaches. Farmers provided the land and in some cases,labor. Researchers provided all other inputs, and made all decisions on inputs (quantity, time andmethod of application) without consulting with farmers. In contrast the baby trials were designed byresearchers but managed by farmers. From the full complement of treatments in the mother trial,farmers hosting baby trials selected the technology options they wanted to experiment with. In nearlyall instances researchers provided seed and fertilizer, as well as advice on inputs to be applied. Farmerswere free to implement their own management practices but most of them followed the advice ofresearchers. Thus, all the baby trials that were monitored followed a prescribed set of technologies andmanagement practices.

8

SMIP/TSBF approach. In the SMIP trials, farmers decided whether to host a research-managed or afarmer-managed trial or both, and which treatments they wanted to experiment with. Researchersdesigned the researcher-managed trials and made all decisions on quantities, when, and how to applyinputs. In the farmer-managed trials, researchers provided seed and facilitated access to fertilizer, butfarmers made the decisions on input application. Researchers gave advice on trial management butfarmers were free to manage their trials as they wanted. Farmers who implemented farmer-managedtrials were aware of what inputs were being used in the researcher-managed trials (this was discussedat the village meetings), but many changed the level, timing, and method of applying inputs to suittheir circumstances.

In the TSBF trials farmers similarly decided whether they wanted to host a researcher-led trial ora farmer-led trial, and what treatments they wanted to test. Researchers designed the researcher-managed trials and made all decisions on inputs, while TSBF technicians and enumerators supervisedand managed the trials. Farmers provided land and labor at cost. TSBF provided all other inputs. In thefarmer-managed trials, TSBF provided inputs including seed and fertilizer, and advice on crop andmanure management. Technicians assisted farmers with plot lay-out. Farmers decided on methods andtiming of input application and overall management of their trials.

Implementation of experimentsImplementation of experimentsImplementation of experimentsImplementation of experimentsImplementation of experiments

Mother-baby approach. The treatments were best-bet technology options designed by researchers.Enumerators collected data and monitored the trials while research technicians supervised the trials.The mother trials, run by research technicians, contained the full set of technology options, whilefarmers planted subsets of these technologies in their baby trials. In some cases enumerators helpedfarmers lay out their trial plots.

SMIP/TSBF approach. Treatments in the SMIP research-managed trials were best-bet optionsdesigned by researchers using information from the diagnostic baseline survey, a cropping systemssimulation model, and researchers' and farmers' prior experience in other areas. In the farmer-led trials,farmers tested either the full set or a subset of technology options that they selected from theresearcher-managed trials, but were free to modify the treatments based on their prior experience.

In the TSBF experiments, treatments in the researcher-managed and farmer-managed trials weredifferent. The former were based on researchers' experiences in other areas. The latter were acombination of treatments based on researchers' experience and modifications arising from the issuesfarmers raised at the village meetings.


Mother-baby approach. Enumerators hired by the different organizations monitored trial plots andcollected data on both the mother and baby trials, although the mother trial was more closelymonitored. All the institutions implementing on-farm trials collected agronomic data (cropemergence, weed cover, crop stand, weeding dates) and farmers' perceptions of the treatments. Inaddition, CIMMYT researchers at the Zimbabwe and Malawi sites collected data on labor use for theentire trial rather than individual treatments. In the ICRISAT-led trials in Malawi and Zimbabwe andthe CIMMYT-led trials in Malawi, enumerators collected data on both the mother and baby trials. Incontrast, the CIMMYT-led trials in Zimbabwe encouraged farmers to collect data on their baby trials.While this helped build farmers' capacity to monitor their experiments, farmers often recorded onlythe dates of key activities such as planting, weeding, germination, and fertilizer application. Most

9

farmers reported that the biophysical data sought by researchers were too complex to monitor. Mostfarmers found data collection useful because they could compare crop performance on trial and non-trial plots. For the most part, the same type of data was collected from mother and baby trials. The datawere passed on to researchers for analysis, but many farmers did not know what the researchers didwith the data.

At each case study site researchers and extension staff organized field days to demonstrate theperformance of the new technologies in the mother trials and in selected baby trials – selectedprimarily by technicians and enumerators. The field days involved trial farmers as well as others fromsurrounding villages. Research and extension staff described the on-farm experiments, led thesubsequent discussion, and answered farmers' questions. These field days provided an opportunity forfarmers hosting baby trials to compare the performance of researcher-managed mother trials with theirown experiments. In addition to field days, CIMMYT researchers in Zimbabwe organized mid-seasonevaluation meetings with farmers hosting the baby trials, to discuss trial performance. These meetingsprovided an opportunity for farmers to make an early assessment of the technologies, and discusspossible adaptations they might want to make in subsequent trials. In the CIMMYT-led trials at onesite in Zimbabwe, researchers encouraged farmer-led field days in conjunction with local extensionstaff, with farmers selecting sites and leading the discussions. In Malawi, ICRISAT researchers usedquantitative evaluation techniques such as matrix and pairwise ranking to obtain farmers' assessmentof the relative strengths and weaknesses of technologies in the baby trials.

SMIP/TSBF approach. In the SMIP trials, enumerators collected agronomic data (including inputapplication) and feedback from farmers' observation of the trials. Similar data were collected in theresearcher-managed and farmer-managed trials. Where possible, farmers were encouraged to collectthe data themselves, using trial books provided by the project. TSBF also used enumerators to collectsimilar data from both researcher-managed and farmer-managed trials. Researcher-managed trialswere monitored more closely than farmer-managed trials. Data collected by farmers were often passedon to researchers.

Both SMIP and TSBF used field days to involve farmers in the research process. SMIP organizedfield days, jointly with national extension staff and farmers' organizations, to demonstrate theperformance of the technologies to other farmers and allow farmers to compare the different optionsbeing tested. Field days were held at both researcher-managed and farmer-managed sites. In theresearcher-managed trials the farmer hosting the trial explained the experiments to other farmers, andresponded to questions with assistance from researchers. In the farmer-managed trials a group offarmers in the community visited all the farmer-managed trials and selected sites for the field days.The farmer hosting the farmer-managed trial explained the trial to other farmers and researchers andled the discussions. TSBF held field days at researcher-managed sites only. Researchers explained thetrials and led the discussions. Farmers hosting the researcher-managed trial explained their level ofinvolvement, their observations, and the problems they faced. Farmers who hosted farmer-managedtrials also explained their experiments and shared their experiences.

SMIP organized end-of-season evaluations at which farmers evaluated researcher-managed andfarmer-managed trials and technology options using matrix scoring and ranking. Farmers developedtheir own evaluation criteria and scored all treatments within trial types according to these criteria.They then ranked treatments overall (within the experiment) and made recommendations for otherfarmers.

TSBF also organized a farmer feedback meeting after the trials were harvested and the dataanalyzed. At these meetings farmer and researchers discussed results from researcher-managed trials;farmers who hosted trials presented their results to other farmers with the assistance of researchers.

10

There was significant variability in the results from farmer-managed trials. Such variability highlightsthe need for researchers to provide farmers with more information on the analysis of trial results.


At all sites where researcher-managed and farmer-managed trials were conducted simultaneously,farmers were encouraged to collect data, monitor, and report on their experiments. This helpedstrengthen their capacity to conduct their own experiments.

Farmer-led with research inputFarmer-led with research inputFarmer-led with research inputFarmer-led with research inputFarmer-led with research input

This approach aims to empower farmers to develop their own solutions to agricultural problems. Thegoal is to encourage the development of farmer leadership within the community, for example throughspecific training exercises such as Training for Transformation. Three types of farmer-ledexperimentation were implemented. ICRISAT and national extension staff in Malawi implementedtrials that were linked to the mother-baby trials; the NGO Concern Universal and ICRISATimplemented farmer empowerment trials in Malawi; and CIMMYT in collaboration with nationalresearch (DR&SS) and extension (AGRITEX) facilitated farmer experimentation in Zimbabwe.


Participatory problem analysis was used as a diagnostic tool in the CIMMYT/DR&SS/AGRITEXtrials in Zimbabwe and the Concern Universal/ICRISAT trials in Malawi. CIMMYT/DR&SS/AGRITEX used resource maps drawn jointly by farmers and enumerators to depict resourceavailability and use, resource flows, product flows, and market relationships. These maps were used astools to identify problems and discuss technology options with farmers. Concern Universal/ICRISATused PRA tools including social mapping, transect walks, and semi-structured interviews to assessfarmers' resources and opportunities for resolving identified production constraints and enhancingincome generation. There was no participatory problem diagnosis at the ICRISAT/national extensioninitiative in Malawi. Rather, researchers relied on earlier diagnostic work on crop productionconstraints in the area. The trials were, however, directly linked to the mother-baby trial design inwhich researchers had formulated hypotheses on priority constraints in the area.

Selecting and prioritizing technology options for experimentationSelecting and prioritizing technology options for experimentationSelecting and prioritizing technology options for experimentationSelecting and prioritizing technology options for experimentationSelecting and prioritizing technology options for experimentation

One objective of the CIMMYT/DR&SS/AGRITEX work was to support farmers to identifytechnology options and conduct their own experiments. Technologies were chosen forexperimentation after discussion between farmers, researchers, and extension in a series of villagemeetings. During these meetings farmers and researchers suggested technology options they had priorexperience with, or which had performed well elsewhere. The final selection of treatments was donethrough negotiation between farmers, researchers, and extension staff.

The PRA implemented by Concern Universal used problem and objective tree analysis forparticipatory problem diagnosis and identification of technology options to address key problems.Farmers participated in problem diagnosis, but their views were not adequately represented in theselection of technology options for experimentation – the experiments conducted on-farm did notinclude several technology options identified in the PRA. Rather, they focused on best-bet optionsidentified by researchers as well as options that reflected Concern Universal's wider development

11

objectives in the area. Discussions with farmers suggested they were interested in testing sometechnology options that were not included in the trials.

In the ICRISAT/national extension work in Malawi, farmers were encouraged to conduct theirown experiments based on their observation of the performance of best-bet options in the mothertrials. Farmers selected the treatments they wanted to try on their own; ICRISAT facilitated farmerexperimentation by providing seed and technical information, and monitoring the experiments.

Planning stagePlanning stagePlanning stagePlanning stagePlanning stage

In all the farmer-led experimentation, farmers managed the trials and made decisions on the quantity,timing, and method of application of inputs. Research and extension staff provided technical advice ontechnology options, ICRISAT and CIMMYT supplied seed for the experiments, and farmers providedall other inputs. In Zimbabwe, CIMMYT suggested a split plot design, which they felt made it easier tocompare farmer practice with the technologies being tested. In both the ICRISAT initiatives inMalawi, farmers decided how they wanted to plant the experiments. Some farmers closely followedthe design of the baby trials while others modified their experiments to suit their circumstances.

Implementation of experimentsImplementation of experimentsImplementation of experimentsImplementation of experimentsImplementation of experiments

In the CIMMYT/DR&SS/AGRITEX trials, farmers planted the treatments they had negotiated withresearchers. Farmers did not change the trial plan or treatments but were free to decide whatmanagement practice to use. At the Concern Universal/ICRISAT sites in Malawi farmers planted thebest-bet options suggested by researchers. Most followed the advice offered by researchers and did notmodify their experiments. In the ICRISAT/national extension trials in Malawi, some farmers plantedthe treatments as they observed them in the baby trials but others modified the treatments to suit theircircumstances.

Enumerators employed by CIMMYT and ICRISAT monitored the farmer-led experiments at allsites. Farmers collected data on dates of operations such as planting and weeding. This data was givento the enumerators but most farmers had no idea what the researchers did with it. Some farmersreported that they have not had an opportunity to discuss the data with researchers. ICRISATresearchers did not collect data on their farmer-led experimentation.


At all sites, farmer-led experimentation was monitored and supervised less intensively thanresearcher-managed trials. Researchers made informal visits to the trials to interact with farmers,discuss trial performance, and get insights into farmers' observations.


The emphasis in all farmer-led research was to facilitate farmer experimentation. Farmers in theCIMMYT/DR&SS/AGRITEX initiative in Zimbabwe were encouraged to collect data and monitortheir experiments. In contrast farmers in both ICRISAT initiatives in Malawi did not collect dataalthough they were encouraged to informally observe their experiments and note their observations.Concern Universal conducted Training for Transformation for farmers who hosted trials. Theobjective of the training was to build capacity and empower farmers to make and manage decisions intheir communities. However, the training was conducted after farmers had planted their experiments,

12

so it is unclear how and to what extent it influenced their decision making processes or communicationwith researchers.

Guidelines for FPR: Key IssuesGuidelines for FPR: Key IssuesGuidelines for FPR: Key IssuesGuidelines for FPR: Key IssuesGuidelines for FPR: Key IssuesThe primary objective of this study was to determine whether there were sufficient differences in FPRmethods across the case study sites to allow a comparison of the approaches and draw guidelines forplanning and designing future projects, particularly as they relate to developing technologies relevantto women farmers.

Characterizing FPR methodsCharacterizing FPR methodsCharacterizing FPR methodsCharacterizing FPR methodsCharacterizing FPR methods

Researchers were requested to characterize the participatory research method they were implementingat each case study site. The findings from this study suggest that researchers correctly characterizedthe research methods they were using into the three broad categories: traditional research-led,research-led with farmer input, and farmer-led with research input. Nonetheless, the study suggeststhat when these methods are characterized in the context of farmer participation, they could best bedepicted not as distinct methods, but rather as a continuum from less farmer participation towardfarmer empowerment, with the largest degree of farmer participation in farmer-led experimentationand the least in traditional research-led approaches. To make the comparison manageable, wesummarized the FPR methods used in terms of the extent of farmer participation or farmer/researcherinteraction at different stages in the research process.

Differences in FPR methods across sitesDifferences in FPR methods across sitesDifferences in FPR methods across sitesDifferences in FPR methods across sitesDifferences in FPR methods across sites

The study suggests that differences in FPR methods across sites were sufficient to allow comparison ofthe different methods. Defining FPR in the context of farmer participation, the study identifiedimportant differences in the ways farmers interacted with researchers at different stages in the researchprocess. All the research programs made efforts to involve farmers in diagnostic activities. Someprograms, such as the traditional research-led approaches and the mother-baby trials in Zimbabwe, didnot conduct specific participatory problem diagnosis. Baseline surveys, in which farmers responded toquestions designed by researchers, were used to diagnose problems at some sites in Malawi andZimbabwe. While some of these surveys were useful in characterizing farmers' production practicesand diagnosing key productivity constraints, there were concerns, especially at the Malawi sites, thatthe survey results were not produced in a timely manner for use in planning future research activities.PRA techniques and tools were used for participatory problem diagnosis at the Concern Universal sitein Malawi and the CIMMYT site in Zimbabwe. These techniques resulted in timely reports that wereuseful in planning subsequent research, but there is a possibility that the experiments eventuallyconducted do not fully reflect the PRA results. PRA type activities also require greater investment oftime by researchers and extension staff. Thus, important decisions must be made in determining whatproportion of total resources to allocate to such activities, and the cost effectiveness of these resourceallocation decisions.

All the research programs used village meetings to elicit farmer participation in diagnosticactivities. However, these meetings were used for different purposes, which also had differentimplications for the ways in which farmers participated. In the traditional research-led and mother-baby trial approaches it appeared that researchers used village meetings to inform farmers about thehypothesized productivity constraint, the experiments that were planned to address key problems, the

13

roles of farmers and researchers, and the manner in which the research program was to be managedand organized. Farmer participation in these meetings can be described as passive – they askedquestions or sought clarifications, but it was not clear how and to what extent their ideas influenced theresearch process. In contrast SMIP and TSBF used village meetings to gain a better understanding offarmers' constraints and priorities, and identify issues that influenced research planning. In these casesfarmers participated as active decision makers, and technology options were selected for on-farmtesting through a process of negotiation between farmers and researchers.

In the planning stage researchers had the greatest control in the traditional researcher-managedapproaches, typically making all decisions about how experiments were to be conducted and withwhat inputs. Farmers were more involved in decisions about where the experiments were plantedbecause they provided land and decided which farmers would conduct the experiments. This situationwas similar in researcher-led mother trials. In the baby trials, farmers were expected to make planningdecisions. However, the manner of implementation suggested that researchers had great control overhow the trials were planted, presumably because the baby trials were meant to be subsets of the mothertrials where researchers had complete control. In addition, researchers largely determined what inputsfarmers used in the baby trials because farmers were advised to use the same plot size as in the mothertrials, and because researchers provided all other inputs besides land and labor. The researcher-ledtrials with farmer input, implemented by SMIP and TSBF, used similar planning arrangements as thetraditional research-led approaches, but farmers were more involved in making decision on how,where, and with what inputs, especially in cases where researchers did not provide the inputs.

In the farmer-led approaches, researchers had the greatest influence in planning the experiments inthe CIMMYT/DR&SS/AGRITEX trials in Zimbabwe because of the close interaction betweenresearchers and farmers in designing the experiments during the diagnostic stage. However, across allsites farmers seemed to have the largest control over the "how, where, and who" decisions in planningexperiments. Farmers had control over the "with what" decisions in cases where they provided theinputs. In contrast, researchers had control over these decisions in cases where they provided farmerswith inputs to conduct experiments.

Field days were used to assess results from the on-farm trials at all research sites, but there wereimportant differences in how farmers participated in field days. In general there were two broad typesof field day. In one type, researchers and extension staff used field days to demonstrate to farmers thepotential benefits of the technologies being tested. Researchers explained the trials, led thediscussions, and answered farmers' questions. This was done in the traditional research-led approachin Zimbabwe and the mother-baby trials in Malawi and Zimbabwe. In the other type of field dayfarmers hosting experiments (both researcher-managed and farmer-managed) explained the trials, ledthe discussions, and responded to questions, sometimes with assistance from researchers. This type offield day was found in the SMIP and TSBF researcher-led trials and the CIMMYT/DR&SS/AGRITEX farmer-led trials in Zimbabwe. In addition to field days, CIMMYT mother-baby trials inZimbabwe used mid-season evaluation meetings with farmers to discuss trial performance, whileSMIP and TSBF used end-of-season evaluations to discuss trial results and set priorities for futureexperimentation. Besides these organized meetings with farmers, researchers in mother-baby trials atall sites in Malawi and Zimbabwe, the SMIP trials in Zimbabwe, and the CIMMYT/DR&SS/AGRITEX farmer-led trials in Zimbabwe, provided opportunities for farmers to note their commentson trial performance in trial books. Researchers also interacted informally with farmers during fieldvisits. Besides the SMIP project, other forms of farmer participation in assessing research results –tours or exchange visits to research areas outside the community and other experiment stations or trialdemonstration sites – were not emphasized in any of the research approaches.

14

Building the capacity of farmers to conduct their own experiments appeared to be limited toefforts to encourage farmers to monitor their trials and collect data. Farmers were collecting datathemselves in CIMMYT's mother-baby trials and farmer-led trials, and the SMIP farmer-managed trialin Zimbabwe. In most cases both enumerators and farmers were collecting the same data. Involvingfarmers in monitoring trials and recording results might enhance farmers' capacity to conduct theirown experiments. But there are concerns whether this could be achieved with the current emphasis onagronomic data rather than on meaningful variables which farmers use in making decisions ontechnology options; for example intensity of resource use (labor, land, capital) and relative returns tothe resources they invest. Across all sites there was also less emphasis on enhancing farmers' ability todesign their own trials or set up trial plots on their own. Similarly, researchers across all sites did notcollect data or consistently monitor farmers' own experiments. This suggests lost opportunities forstimulating farmer experimentation or broadening the research agenda to work on solutions thatfarmers have taken the lead on.

TTTTTargeting women farmersargeting women farmersargeting women farmersargeting women farmersargeting women farmers

A primary focus was to develop technologies relevant to women farmers. Yet none of the trialsspecifically targeted women or explicitly incorporated gender issues in the research process. Severalcriteria were used to select farmers for participation in on-farm trials, but almost none targeted women.In part, this omission reflected the lack of procedures for targeting women in stratification of farmersor as targets for empowerment. Thus, in practice the participation of women farmers inexperimentation was largely ad-hoc and turned out to be the most difficult aspect of participatoryresearch.

Objectives for different types of trialsObjectives for different types of trialsObjectives for different types of trialsObjectives for different types of trialsObjectives for different types of trials

Comparison of FPR methods was made difficult because the project did not formulate clear objectives,across locations, for the different trials. The project document stated one main objective: to comparethe effectiveness of farmer-led versus researcher-led FPR methods in the development anddissemination of technology options. Nonetheless, considering the potential for variability inbiophysical and socioeconomic parameters within and across sites as well as the effects of farmermanagement on performance/assessment of technology options, it is doubtful whether one generalobjective can be applied to different types of on-farm trials. The potential for variability in on-farmtrial results implies the need to formulate clear justifications and objectives for conducting differenttypes of research, and spell out the expected role of farmers in achieving the stated objectives. Thechallenge for researchers is to effectively integrate different types of research to develop technologiesthat are practical and profitable.

ConclusionsConclusionsConclusionsConclusionsConclusionsThis study used a case study approach to compare different types of farmer-participatory researchbeing implemented in Malawi and Zimbabwe. The study used a framework for comparing differentFPR methods based on stages in the research process and the level of farmer/researcher participationin decision making. The research process was divided into three stages: diagnostic; planning ofexperimentation; testing and adapting. On-farm trials were implemented at case study sites organizedaround trial villages. Researchers characterized their work at each site as traditional research-led,research-led with farmer input, and farmer-led with or without research input. It was hypothesized that

15

the largest degree of researcher control would be in a traditional trial approach, with controldiminishing in favor of farmer control as the research evolved towards a farmer empowermentapproach.

The study found that researchers correctly characterized their research method in broad terms.However, the practice of FPR at different sites in the context of farmers' participation suggested that,rather then being distinct or separate approaches, these methods represent a continuum from lessfarmer participation toward farmer empowerment; with the largest degree of farmer participationin farmer experimentation and the least in traditional research-led approaches. The study founddifferences in diagnostic activities, planning of experiments, assessment of trial results, andstrengthening of farmers' capacity to conduct their own experiments. These differences would permita comparison of the different approaches. The study also suggests that researchers need to pay greaterattention to procedures for targeting women farmers, integrating gender issues into the researchprocess, and formulating clear justifications and objectives for different types of on-farm experiments.

AcknowledgmentsAcknowledgmentsAcknowledgmentsAcknowledgmentsAcknowledgmentsThis study arises from an ICRISAT project Will women farmers invest in improving their soil fertilitymanagement? Participatory experimentation in a risky environment. I am grateful to Sieglinde Snappfor her encouragement and support for the study. Several people contributed to the preparation andimplementation of field work and I particularly thank Walter Chivasa, Doubt Gumbonzvanda, PatriciaMasikate, Ransom Chimbaranga (late), Bernard Matsikachando, Jacob Nyirongo, and all theenumerators at the field sites, whose inputs were invaluable to the success of the study. Special thanksare due to the farmers who participated in the informal surveys and discussions. Kit Vaughan, BernardKamanga, Herbert Murwira, Jacob Mapemba, and Lucia Muza provided valuable insights into theirFPR work that formed the basis for this study. Additional insights for the paper came from participantsat two workshops in Bulawayo. I am grateful to Hugo de Groote, Geoff Heinrich, David Rohrbach, andSteve Twomlow for comments on earlier drafts of this paper. The support of the UK Department forInternational Development in funding the project is gratefully acknowledged. The author, however, issolely responsible for the views expressed in this paper.

ReferencesReferencesReferencesReferencesReferencesLilja, N., Lilja, N., Lilja, N., Lilja, N., Lilja, N., and AshbyAshbyAshbyAshbyAshby, J.A., J.A., J.A., J.A., J.A. 2000. Types of participatory research based on locus of decision-making.PRGA Working Paper No. 6. Washington DC, USA: CGIAR Systemwide Program on ParticipatoryResearch and Gender Analysis for Technology Development and Institutional Innovation.

Okali, C., Sumberg, J.,Okali, C., Sumberg, J.,Okali, C., Sumberg, J.,Okali, C., Sumberg, J.,Okali, C., Sumberg, J., and Farrington, J.Farrington, J.Farrington, J.Farrington, J.Farrington, J. 1994. Farmer participatory research: rhetoric and reality.London, UK: Intermediate Technology Group.

Snapp, S.S. Snapp, S.S. Snapp, S.S. Snapp, S.S. Snapp, S.S. 1999. Farmer-participatory approaches for working with women farmers. Pages 28-32 inImproving soil management options for women farmers in Malawi and Zimbabwe (Snapp, S.S., ed).Proceedings of a regional workshop "Will women farmers invest in improving their soil fertilitymanagement? Participatory experimentation in a risky environment", 17-19 May, 1999. PO Box 776,Bulawayo, Zimbabwe: International Crops Research Institute for the Semi-Arid Tropics.

16

Annex 1. Case study sites, FPR methodologyAnnex 1. Case study sites, FPR methodologyAnnex 1. Case study sites, FPR methodologyAnnex 1. Case study sites, FPR methodologyAnnex 1. Case study sites, FPR methodology, and lead, and lead, and lead, and lead, and leadorganization(s)organization(s)organization(s)organization(s)organization(s)

Location Village FPR method Lead organization(s)

MalawiMalawiMalawiMalawiMalawiDedza Kanti Research led, farmer input ICRISAT/Concern UniversalDedza Kulumeka Farmer led, research input Concern UniversalChisepo Mbinga Research led, farmer input CIMMYT/ICRISATChisepo Santhe Farmer led, research input CIMMYT/ICRISATChisepo Kampenga Research led, farmer input CIMMYT

ZimbabweZimbabweZimbabweZimbabweZimbabweZimuto Mahoto Research led, traditional DR&SSZimuto Mahoto Research led, farmer input DR&SS/ICRISATZimuto Maraire Research led, farmer input CIMMYTZimuto Chikato Farmer led, research input CIMMYTTsholotsho Nembe Research led, farmer input DR&SS/ICRISATTsholotsho Nhalangano Research led, farmer input SMIP and TSBFTsholotsho Nhalangano Farmer led, research input SMIP and TSBFTsholotsho Manzimahle Research led, farmer input SMIPTsholotsho Manzimahle Farmer led, research input SMIP

Annex 2. Checklist for discussion with farmersAnnex 2. Checklist for discussion with farmersAnnex 2. Checklist for discussion with farmersAnnex 2. Checklist for discussion with farmersAnnex 2. Checklist for discussion with farmers

Level of historical interventionWhoSince whenWhat issues

Selection of trial farmers

Who selectsWhy selected

Modes of participation

Identification of constraints and opportunities

Who identifies subject matter for investigationHow were these identified – instruments used

Identification of ideas/options

Trial design

Who identifies technology options to be testedHow are priorities setWho determines type of trial and choice of treatmentsWho designs trials

... continued

17

Annex 2.Annex 2.Annex 2.Annex 2.Annex 2. Continued.

Testing/adapting options

Trial management

Who decides on level of inputs (quantity applied, when, how)Can farmer change decisions on inputs (consult? who?)Who supplies inputs

Data collection and analysis

Who collects dataTypes of data collected, variables monitoredWho analyzes dataHow are results evaluatedHow are data interpretedHow are results disseminatedHow is information shared, with whom

Intensity of interaction

Mode of interaction – researchers, technicians, extension, farmerFrequency of interactionInvolvement with on-station and on-farm experimentation

Farmer knowledge and perceptions

Knowledge about objectives of trialsPerception of treatmentsEstablish trials on their own (understanding the research process)

Farmer experimentation

What experiments conducted

Motivation for conducting experimentsMonitor farmer experimentationMechanism for feedbackInfluence research agenda (how and to what extent)

18

Annex 3. Types of trial implemented at case study sitesTYPE OF TRIAL: Traditional research-led

Research process Researcher only, Researcher with Researcher and Farmer with Farmer withoutno farmer passive farmer farmer jointly researcher researcher involvementinvolvement involvement (negotiated involvement

process)

Diagnose problem √

Identify opportunities √

Set priorities √

Identify options √

Planning experimentsHow √Where √Who √With what √

Conducting √experimentation

Assessing results √

TrainingPlot layoutReplicationMonitoring, √data collection

19

TYPE OF TRIAL: Researcher-led – mother-baby trial design


process)



Set priorities √


Planning experimentsHow √Where √ √Who √ √With what √

Conducting √ √experimentation

Assessing results √ √ √

TrainingPlot layoutReplicationMonitoring, √ √data collection

20

TYPE OF TRIAL: Researcher-led – simultaneous research-managed and farmer-managed research


process)



Set priorities √


Planning experimentsHow √ √Where √ √Who √With what √ √


Assessing results √ √

TrainingPlot layoutReplicationMonitoring, √data collection

21

TYPE OF TRIAL: Farmer-led research – farmer experimentation linked to mother-baby trials


process)



Set priorities √





TrainingPlot layoutReplicationMonitoring,data collection

22

TYPE OF TRIAL: Farmer-led research – facilitating farmer experimentation


process)



Set priorities √





TrainingPlot layout √Replication √ √Monitoring, √data collection

23

TYPE OF TRIAL: Farmer-led research – farmer empowerment


process)



Set priorities √





TrainingPlot layout √ReplicationMonitoring,data collection

About ICRISAT

The semi-arid tropics (SAT) encompasses parts of 48 developing countries including most of India,parts of southeast Asia, a swathe across sub-Saharan Africa, much of southern and eastern Africa, andparts of Latin America. Many of these countries are among the poorest in the world. Approximatelyone-sixth of the world’s population lives in the SAT, which is typified by unpredictable weather,limited and erratic rainfall, and nutrient-poor soils.

ICRISAT’s mandate crops are sorghum, pearl millet, finger millet, chickpea, pigeonpea, andgroundnut; these six crops are vital to life for the ever-increasing populations of the semi-arid tropics.ICRISAT’s mission is to conduct research which can lead to enhanced sustainable production ofthese crops and to improved management of the limited natural resources of the SAT. ICRISATcommunicates information on technologies as they are developed through workshops, networks,training, library services, and publishing.

ICRISAT was established in 1972. It is one of 16 nonprofit, research and training centers fundedthrough the Consultative Group on International Agricultural Research (CGIAR). The CGIAR is aninformal association of approximately 50 public and private sector donors; it is co-sponsored by theFood and Agriculture Organization of the United Nations (FAO), the United Nations DevelopmentProgramme (UNDP), the United Nations Environment Programme (UNEP), and the World Bank.

AbstractFarmer-participatory research (FPR) can help improve the effectiveness of technology development, raiseadoption rates, and increase the payoff to agricultural research. However, there is wide diversity of opinion aboutthe scope and nature of farmer participation in the implementation of FPR. This study compared different FPRapproaches being used by various organizations at six on-farm sites in Malawi and Zimbabwe, with a view todeveloping guidelines for future work, particularly in developing technologies targeted at women farmers. Thestudy found differences between the various approaches in diagnostic activities, planning of experiments,assessment of results, and strengthening of farmers' capacity to conduct their own experiments. Overall, however,the various methods were not completely distinct, but represented a continuum from traditional research-led, toresearcher-led with farmer input, to farmer-led with research input. The results indicated that when there are noclear procedures to directly target women farmers they tend to be under-represented, and gender issues are notsufficiently integrated into the research process. PRA techniques were also found to be useful in planningexperiments, but require considerable investment of time by researchers and extension staff.

Researchers therefore need to formulate clear objectives for conducting different types ofon-farm experiments, ensure that farmers understand their roles in each case, and finally integrate different typesof research to develop technologies that are practical and profitable.

RésuméComparaison des méthodologies de recherche menée avec la participation des paysans : Etudes de cas auMalawi et au Zimbabwe. Les recherches menées avec la participation des paysans (FPR) peuvent améliorer lamise au point des technologies, augmenter le taux d’adoption et accroître les retombées positives de la rechercheagricole. Il existe cependant, une large diversité d’opinions sur la portée et la nature de la participation despaysans à la mise en œuvre de la recherche participative. La présente étude a comparé les différentes approchesen matière de recherche participative adoptées par différentes organisations dans six sites en milieu paysan auMalawi et au Zimbabwe, en vue d’élaborer des directives pour les travaux futurs, en particulier pour la mise aupoint de technologies destinées aux femmes en milieu rural. L’étude a trouvé qu’il y a des différences entre lesdiverses approches adoptées dans les activités de diagnostic, de planification des expériences, d’évaluation desrésultats et de renforcement de la capacité des paysans à réaliser leurs propres expériences. En général, cesdifférentes méthodes ne se distinguaient pas totalement les unes des autres, mais formaient un continuum qui vades méthodes de recherche sans participation des paysans, ensuite les recherches faites par les chercheurs avec laparticipation des paysans et à la fin la recherche faite par les paysans en collaboration avec les chercheurs. Lesrésultats de l’étude indiquent que lorsqu’il n’y a pas de procédures claires permettant de cibler directement lesfemmes productrices, ces dernières ont tendance à être sous-représentées et les questions de genre ne sont passuffisamment prises en compte dans le processus de recherche. L’étude a également trouvé que les techniques deFPR sont utiles pour la planification des expériences, mais exigent un important investissement de temps de lapart des chercheurs et du personnel de vulgarisation. Les chercheurs doivent donc définir des objectifs clairs pourla conduite de différents types d’expériences en milieu paysan; s’assurer que les paysans comprennent leur rôledans chaque cas et combiner différents types de recherche pour mettre au point des technologies pratiques etrentables.

Citation:Citation:Citation:Citation:Citation: Freeman, H.A. 2001. Comparison of farmer-participatory research methodologies: case studies inMalawi and Zimbabwe. Working Paper Series no. 10. PO Box 39063, Nairobi, Kenya: Socioeconomics andPolicy Program, International Crops Research Institute for the Semi-Arid Tropics. 28 pp.

Titles in the Working Paper Series aim to disseminate information and stimulate feedback from the scientificcommunity. A large number of similar as well as informal publications have been distributed by ICRISAT’sSocioeconomics and Policy Program, under various titles: Progress Reports, Discussion Papers, and OccasionalPapers.

H A Freeman

International Crops Research Institute for the Semi-Arid Tropics



1030–2001

Working Paper Series no. 10

Socioeconomics and Policy Program

Comparison of Farmer-ParticipatoryResearch Methodologies:

Case Studies in Malawi and Zimbabwe

Documents

Comparison of Farmer-Participatory Research Methodologies